Distributed Dual-rate Consensus Predictive Control of Looper Tension System in Hot Rolling Mills

Abstract

This paper considers a dual-rate distributed predictive control strategy for the looper tension system in hot rolling mills, which is a typical multi-agent system with directed communication topology. First, we establish an interconnected model for looper tension control system and the disturbances from the neighbors are considered effectively. Second, the consensus control protocol is developed based on the proposed control strategy to improve the robustness and stability of the multi-agent, and the sufficient conditions for consensus are developed. We update and implement all the agent controllers sequentially in one output sampling period and begin a new cycle at the next sampling instant, which leads the multi-agent control system is of fast control updating rate and slow output sampling rate. The control inputs of neighbors can be obtained to compensate the coupling effects, and the cooperation of controllers are improved. Finally, simulation results verify the proposed control strategy and corresponding results.